Tall Dads, Shorter Kids, and the Insoles That Cured Nothing: A Field Guide to Getting Better on Your Own

Somewhere in America right now, a person is writing a five-star review for a product that did nothing.

Maybe it's the $200 orthotic insoles. Maybe it's the copper bracelet, the elimination diet, the supplement with a name that sounds like a Marvel villain. The pain got better. The purchase came first. The reviewer connects the dots — A before B, therefore A caused B.

The problem is that B was already on its way. Back pain, allergy flare-ups, knee twinges, even depressive episodes tend to be worst right before people seek treatment, and conditions at their worst have nowhere to go but toward average. Statisticians call it regression to the mean. Everyone else calls it a miracle.

The Victorian Problem That Started It All

The concept has a surprisingly wholesome origin story. In the 1880s, Francis Galton — a Victorian polymath with, it must be said, some truly terrible ideas about eugenics — was measuring families. He noticed something odd: the children of exceptionally tall parents were, on average, shorter than their parents. And the children of short parents were taller.

He called it "regression towards mediocrity," which is honestly a better name than the one we ended up with.

Galton's data showed that height doesn't march ever-upward through generations. Extremes get diluted. Very tall parents tend to have tall-ish kids — above average, but pulled back toward the middle. This wasn't genetics working against itself. It was what happens any time you measure something twice and the second measurement is partly luck.

The key insight: whenever a variable has some random component, extreme values are loaded with good (or bad) luck. And luck doesn't repeat. The tallest dad in the room probably caught a few statistical breaks. His kid inherits the genes, not the breaks.

Why Your Worst Day Is the Worst Possible Baseline

Here's the part that matters for your wallet.

You don't buy insoles on a random Tuesday. You buy them after a brutal week — after the flight, after the 5K, after you spent a weekend moving furniture. You seek help when the pain is at its peak. And a peak, by definition, is an extreme value loaded with bad luck.

If you measured your pain every day for a year and drew a chart, you'd see it wandering up and down. The worst point on that chart is almost guaranteed to be followed by improvement — not because the chart knows you bought something, but because extremes don't persist. Randomness pulls you back to your personal average.

This is why every remedy works. Acupuncture, crystal healing, essential oils, a new mattress, a sports drink endorsed by a quarterback — anything you try at your low point gets credited with a recovery that was already scheduled. The worse you feel before treatment, the more dramatic the "response."

Researchers have known this for decades, which is why real clinical trials use control groups. The placebo group isn't there to test whether sugar pills cure things. It's there to measure how much people improve on their own — and the answer is: a lot.

The Sports Illustrated Cover Jinx, Explained

Regression to the mean doesn't just live in medicine. It haunts sports, and Americans have built entire mythologies around it.

The "Madden curse" and the Sports Illustrated cover jinx work identically: you end up on the cover because you just had a career-best season — an extreme value. Next season, you regress toward your average. Fans see a curse. Statisticians see a bell curve.

The same logic explains the "sophomore slump" for rookie of the year winners, why teams featured in fawning magazine profiles tend to come back to earth, and why the coach who bench-pressed his way to a turnaround rarely repeats it. Success earns attention; attention arrives at the peak; the peak passes. Nobody writes a cover story about regression to the mean, which is a shame, because it would sell about as many copies as the jinx.

The Testimonial Machine

Now layer this on top of modern commerce. Every product page collects reviews, and reviews are written at moments of relief. A weight-loss supplement gets testimonials from people who started a diet at their heaviest. An allergy remedy collects praise from sufferers who tried it during a record pollen count — a count that was going to drop regardless.

The testimonial pipeline is essentially a machine for harvesting regression to the mean and selling it back to you. No one reviews the insoles they bought during an unremarkable week, because they didn't buy insoles during an unremarkable week. Selection bias guarantees the testimonials, and regression guarantees the testimonials are positive.

The fix is boring but reliable: ask what happened to the people who didn't buy the thing. In a fair comparison, the insole-buyers and the do-nothing group improve by nearly the same amount, and the difference between them — not the dramatic before-and-after — is the only number that matters.

How to Spot It in the Wild

You don't need a statistics degree. You need three questions:

Was the starting point extreme? Peak pain, peak weight, peak slump — if the "before" photo is the worst-case scenario, some recovery was coming free of charge.

Did everyone else improve too? If you can't find what happened to people who did nothing, you're being shown half a dataset.

Is the story about the average or the exceptions? Testimonials are cherry-picked by design. One grateful customer tells you nothing; a hundred tell you slightly more than nothing; a controlled comparison tells you everything.

Galton figured this out measuring families in the 1880s, before anyone had heard of a p-value. The math hasn't changed. The testimonials have just gotten better fonts.

The next time something "worked," consider the possibility that you were simply due. It's less satisfying than a miracle — but unlike a miracle, it's true.